Evidence review (PDF) published in March 2015.
Broadband can have a positive impact on the local economy, but the effects are likely to vary across types of firms, workers and areas.
What is broadband?
Broadband is always-on access to an internet service. There are many varieties of broadband technology in use around the world today, offering a range of speeds and modes of access. Broadband is generally considered to be a ‘general purpose technology’ in that it functions across many areas of economic and social life, and it is an enabler of further innovation in those fields.
The rationale: How does broadband deliver growth?
For firms, broadband should allow for efficiencies in production, by lowering costs (for data storage, advertising, or working with suppliers) and enabling innovation (reaching new customers online or employing big data analytics). Those productivity gains could potentially translate into higher wages, and possibly higher levels of employment.
Broadband could also allow for more flexible patterns of work, including working at home or on the move. This flexibility may increase labour force participation, which could in turn raise employment. Broadband may also lower the barriers to starting a business. Increased labour force participation might raise overall employment levels, but that increase in labour supply might depress wages – or lower employment rates.
If broadband makes industries more competitive, some firms will lose staff or face going out of business. Broadband might help accelerate automation, which penalises less skilled workers and those doing routine tasks. It may also have different economic impacts in different types of places.
Governments can provide broadband infrastructure or services, or subsidise connections to existing networks (indirect provision), either through voucher schemes or loans. Governments may also part-fund infrastructure in locations where there is market failure (such as some rural areas) or provide basic funding to incentivise research and development.
Evidence review: What does the evidence say about broadband?
Our review considered more than 1,000 policy evaluations and evidence reviews from the UK and other OECD countries. It found 16 impact evaluations that met our minimum standards.
What the evidence showed:
- Extending broadband to an area can affect firm productivity, number of businesses, and local labour market outcomes (such as employment, income and wages).
- These effects are not always positive, are not necessarily large, and may depend on complementary investments by firms (for example, training workers, or reorganising sales strategy or supply chains to take advantage of faster internet connections).
- Effects can vary across different types of industries and workers with service industries and skilled workers possibly benefiting more than manufacturing industries and unskilled workers.
- The economic effects of broadband tend to be larger in urban areas (or close to urban areas) than in rural areas.
Where there was lack of evidence:
- Most studies look at the effect of broadband provision. Only two studies compare broadband adoption with provision. It is hard to generalise conclusions from these two studies, but they do suggest that the effects of adoption and provision may differ.
- Only two studies look at the effects on profits and sales – one shows a positive effect (for US farms), but the other shows zero effect (for firms in East Yorkshire).
- Only one study looks at the effects on property prices – showing a positive effect on residential property prices.
- We have surprisingly little evaluation evidence of broadband’s impact on working patterns – one study finds that broadband positively affects female labour force participation; another study from 2012, however, finds no net effects on working at home, telecommuting or operating a home-based business.
- We only found three high quality evaluations of specific broadband policies (voucher schemes, direct public provision or public/private partnerships).
- Costs are rarely addressed in the studies reviewed. Only one paper attempts a cost-benefit analysis.
- We have no studies that evaluate SME-targeted voucher schemes.
- More research is required about the relative effects of indirect vs. direct provision (i.e. voucher schemes for services vs. direct investment infrastructure).
- There is a lack of evidence in other areas of internet technology such as the effect of Wi-Fi networks, and fast mobile internet. Future evaluations in this area would greatly improve the evidence base.
- Broadband is a ‘disruptive’ technology that creates winners and losers, and has spillover effects across local boundaries. It is not a silver bullet for local economic development.
- There is evidence that broadband has positive local economic impacts, but some of these may be due to in-migration. Existing residents may not be the biggest beneficiaries.
- Broadband seems to benefit skilled workers more than low- or unskilled workers.
- The effects of adoption and provision may differ. More work needs to be done to understand whether and how to encourage adoption and productive business use.
- Rural areas may need to subsidise broadband provision but the economic benefits of doing so will not be as large as for urban areas.
Toolkits: Advice on designing broadband programmes
In addition to the evidence review, we have three policy design toolkits to help you to make informed decisions when designing broadband interventions. Each toolkit covers a different policy approach. The toolkits consider a broader evidence base than the evidence review.
Case studies: Advice on how to evaluate broadband
Evaluation of the local economic growth effects of broadband can be particularly challenging, as selection is likely to be a much bigger problem for this policy than many others. These selection effects mean that underlying trends must be disentangled from any potential effects of broadband access.
Private sector broadband providers invest only where they expect there to be strong and growing demand. These locations are likely to be already experiencing economic growth and increases in jobs and wages. For this reason, treated areas are almost always likely to be different to untreated areas. Some of these differences will be hard to observe in available data, making it very difficult to construct an appropriate control group. Furthermore, these unobservable differences are not fixed over time, since suppliers have large incentives to respond very quickly to changes in demand, and to be informed and sophisticated in predicting future economic success of locations. For this reason, many evaluations of broadband exploit some source of randomness.
To help improve evaluation of broadband, we have two case studies that give examples of how previous broadband programmes have been evaluated.
Each evaluation case study has met our minimum standard of evidence, which means it (at a minimum) compares what changed for the places, businesses or individuals that benefited from an intervention with what changed over the same time frame for otherwise comparable places, businesses or individuals that didn’t benefit, or that received a different type of intervention. Both broadband case studies use a statistical approach to achieve this comparison. This approach tries to ‘strip out’ the impact of the other factors that could have affected outcomes in both the beneficiary group and the comparator group.
Read more about how to evaluate, and why we think it can be helpful to learn from previous approaches in our how to evaluate guide. You can also read more about different evaluation methods in our scoring guide.